Five Questions Regarding the Australian Share Market for Australian Based Investors.
(This does not examine comparative investments in other markets, commodities or currencies)
1. Is there a possible buying opportunity?
2. Should I buy some now?
3. Can I lose by buying now?
4. Is my reliance on these indicators appropriate for the market we are in?
5. Is the 12 month macro environment/outlook positive or negative?
Is there a possible buying opportunity?
Yes!
Should I buy some shares now?
Yes! The 10 day simple moving average has turned up after a fall of 20% (using the intraday low).
Can I lose by buying now?
Yes! The 10 day SMA is volatile and gets whipsawn, but no market recovery starts without the 10 day SMA turning up. Based on the price pattern after the 1987 crash, you can expect to make a loss from whipsawing until there is a significant major uptrend. That major uptrend might start now in which case you will have no drawdown or there may be say 10 false signals, many involving a loss on the round trip trades, until a major uptrend starts. But, you will not be caught in a future major downtrend if you follow a moving average strategy. You may, however, not do as well as if you just bought in 100% now and held as long as possible.
Is my reliance on these indicators appropriate for the market we are in?
This depends on your age/recovery time, other assets, income outlook but if you are highly concerned about major losses (the Japanese scenario where their market is now down 78% from its all time high in 1990) but also want to participate in any upside, a moving average strategy will protect you from a Japanese scenario but at the likely cost of reduced returns in a sideways or uptrending market (This statement based on analysis by Doug Short: www.dshort.com).
Is the 12 month macro environment/outlook positive or negative?
My opinion is that it is negative. This is based on the unresolved PIIGS debt problems, the introduction of fiscal austerity in US, UK and Australia as well as in the PIIGS and the current trailing earnings in the US being at a never sustained multiple of the 10 year average real earnings. The abandonment of mark-to-market also causes concern. However, central banks are taking extraordinary measures to ensure financial stability and to destroy incentives to hold cash and short and even long term securities. The US 10 year rate is about 2.2% and the S&P 500 earnings yield is about 5%, a ratio rarely sustained for long since 1960, so stocks are quite attractive if earnings merely hold up at current levels. I follow the hypothesis of Richard Koo that this is a balance sheet recession and that to avoid GDP falls the government must take the savings of the private sector and spend them.
Detailed explanation.
I analyse the relative performance of the Australian All Ordinaries Index since 1984 including:
1. returns over 2 mnths and 1, 2, 3, 4 and 5 years
2. whether it is overbought/sold against its 200 day SMA
3. its position against its all time and most recent tops
4. the length and size of its most recent rise or fall against all rises and falls of more than 8% since 1984. (8% is a not quite arbitrary number to eliminate the noise of minor, short term fluctuations).
5. its volatility (persistent high volatility is correlated with major bear markets)
6. the comparative percentage recovery over time from major bottoms
7. the rises and falls between major tops and bottoms and between tops and bottom 3 years apart. (to smooth numbers and avoid comparison only to some extraordinary crash or bubble)
8. What gains/losses have been for 1,2,3,4 and 5 years since the previous tops and bottoms to assist in identifying whether based on past growth rates for those periods at prior tops and bottoms we are likely near a top or bottom (but this analysis will not assist in a long term secular downtrend like Japan as each cyclical top is often lower than the previous cyclical top).
Buy indicators
Present indicators of now being a likely buying opportunity include (based on prices since 1984):
1. The Australian market has only been below its highest point previously by the current 37% for 8% of the trading days since 1984, it has been closer to its previous all time high 92% of trading days since 1984 and has always reached new highs.
2. The length of the current fall to its recent bottom is above both average and median duration of all falls of greater than 8% since 1984
3. The size of the current fall to its recent bottom is above both average and median duration of all falls of greater than 8% since 1984
4. My measure of volatility is in the 96th percentile, so near all time highs
5. Growth over 2 months and 1, 2, 3, 4, and 5 years is at significantly lower than median points being in the bottom 10th, 18th, 18th, 10th, absolute bottom and 2nd percentiles.
6. The market has, since 1984, hit a new record high within 5.5 years, so we might expect to be at 6850 by April 2013, which would be a rise of 60% over the next 20 months (personally I doubt it very much, but the September 1987 high was exceeded in January 2004 in spite of the fall of over 50% which took place in September 1987.
7. The percentage recovery of 37.7% at trading day 625 since the bottom is below average and below all but the 1987 recovery at day 625. Until this recent fall the current recovery had never been below average since day 20. But the 1987 recovery had a sustained period of relapse lower than the current 37.7% recovery after day 625. The 1987 recovery went through a further major downturn and spent only 170 of the next 586 days (the limit of my analysis) above the present level. So there is risk of being whipsawn for a loss if you buy now and get a sell signal later.
8. After the recent falls the market is significantly below its linear trend line on a simple scale and its logarithmic trend line on a logarithmic scale so mean reversion would indicate a likely future uptrend, although timing is always uncertain.
Risk indicators
1. the 30, 50, 100 and 200 smas are all still trending down.
2. there have been no bullish crosses of the 10/30, 10/50, 30/50, 30/100, 50/100 or 50/200 moving averages, so there is a real possibility that you will get whipsawn for a loss if you buy now. It may be too early.
3. You can't totally rule out a Japanese style situation of cyclical losses for another 10 years.
4. The 1987 situation went to a lower level of recovery for about 76% of the next 586 days
5. Periods earlier than 1984 (particularly the post 1929 years) will likely have worse results but are not included in my analysis.
My record
As for my record of profit, I only became an active manager of my funds since the peak of 2007 and I got in too early after missing some parts of the fall and from the top in 2007 I am now slightly ahead of the market in nominal terms.Measurement is complicated by withdrawals to fund retirement. When I add back my cash withdrawals for living expensesI am down approximately 12% from October 2007 compared to the market being down 37%.
My future strategy
I will sell again if there are falls resulting in a downturn of the 10 day moving SMA but not necessarily on the day of the downturn. I will buy more if more trend lines turn up and there are bullish crosses of some of the trendlines I monitor. I have not yet decided exactly what my buy actions will be based on.
Disclosure
I have a significant exposure to the All Ords established within the last 2 weeks (before the 10 day sma turned up but based upon the size of the fall and the percentiles of volatility and growth), but less than 50% of my investment pool, so at present I am almost ambivalent about the direction of the market. I have locked in some outperformace against the All Ords but can take advantage of any major falls to buy in at lower prices, but will regret not having bought more if prices rise and my 10 day SMA remains in an uptrend.
My 10 day SMA is based on the sum of the last 3 days minus the sum of the previous 3 days
Thursday, August 25, 2011
Monday, April 11, 2011
All Ordinaries Market Overview Since 1984 - Looking Forward
Background and Analysis Discussed
Yahoo Finance provides data for the Australian All Ordinaries Index since 1984.
I have used that data to analyse the Australian Market into percentiles of performance. I have then looked at those percentiles at peaks and troughs. I have also looked at what the percentiles of 1, 2, 3, 4 and 5 year growth would be at the 3rd anniversary of the market bottom on 6 March 2009 at different levels of performance over the next 12 months to see what the market would then look like against historical peaks and troughs. I have started some of the analysis since 1984, but some has been constrained because the time performance can oly be calculated at the relevant number of years after the first index value is available.
I have also charted each recovery since from the 1987 bottom in percentage terms to be able to compare the size and speed of recovery in percentage terms by number of trading days since the bottom.
My goal is to provide useful guidance as to the likelihood of future performance, based on the historical recoveries and performance of the Australian All Ordinaries since 1984.
Comparative Speed and Scale of Recovery from March 2009
After 531 trading days, every other recovery since 1984 was lower in percentage terms than today's recovery of 62.1% from the 2009 low. The only market since 1984 which had a higher percentage recovery at any time within 531 days was the recovery from 1992 which had recovered 71.8% by trading day 310, but after 531 days had fallen back to a recovery of only 39.8%.
Looking at it another way, the current recovery is 70 days ahead of the 2003 recovery (the next strongest) and 319 days ahead of the average recovery since the 1987 bottom.
Looking out to day 800 (chosen as it is the point with the widest range of percentage recoveries since 1984 and only170 trading days away from now for this recovery):
1. the 2003 recovery was at 97.4% (best) (an implied future rise of 36.3 points to an index value of 7993 from today's intraday value of 5044, an index rise of 58.4%)
2. the 1987 recovery was at 5.9% (worst) (an implied future fall of 56.2 points to an index value of 3420, an index fall of 32.2%
3 the average of all recoveries was at 60.1% (implying an almost insignificant fall from today).
Of the 5 other recoveries from 20% falls since the 1987 bottom, 3 had ended and had 20% falls by day 531. Only the recoveries from the 1995 and 2003 bottoms went for longer without a 20% fall. The biggest fall in this recovery has been the 11.3% fall of 567.4 points from 5024.1 on 15 April to 4456.7 on 15 July 2010.
Historical Percentile Performance of the Australian Market
The chart below shows the percentiles of performance over rolling 1 to 5 year periods since 1991. It was compiled based on performance to 5 April 2010. From the chart it can be seen that the 5 year performance is between -10% and 50% about 54% of the time. Similarly it can be seen that 1 year performance is less than 0 about 30% of the time.
From this you might assume that at any point in time the likelihood is for positive growth over each of these time horizons, but this ignores that the strongest growth, at least for 1 and 2 year horizons, generally follows a market bottom eg after 2009. From the chart though it seems that if all horizons are in the say lowest 10 percentiles (1 to 10) it is likely a market bottom and if they are all in the highest say 10 percentiles (90 to 100) it is likely a market top. The starting date used (1991) is to ensure all time horizons are looked at from the same starting date, in this case near the bottom of the post 1987 performance for all 5 horizons.
Note that the longer the horizon the more the performance can lag major changes in the index. Note that at present while the All Ords are up 61% from the 2009 bottom it is only reflected in the green 2 year performance with all other horizons showing relatively poor performance.
Percentile Heat Map
We can get a general impression of what are good and bad levels of performance over different time horizons by looking at a percentile heat map. If most performance horizons are in the top 10 percentiles it's probably time to consider rebalancing to cash or bonds, or lengthen duration of bonds, particularly if the yield curve is inverted. On the other hand, if the yield curve is unusually steep and most performance horizons are in the lowest decile then it is time to consider rebalancing toward equities and perhaps, if you have modeled the range of possible negative outcomes and have recent experience, even considering buying call options, geared equities or contracts for difference.
The heat map below tells you that since 1991 a 1 year performance of -14.6% or less is in the bottom 10% of outcomes for 1 year performance and is a likely buying opportunity. Similarly a 5 year performance of 81% or better is better than 90% of results and it is likely a dangerous time to buy or hold.
Recognise however that because of lags in the longer time horizons a 3, 4 or 5 year performance can still be quite good even as the market approaches a bottom and a good 2 year performance now will likely be translated into a good 3 year performance in 12 months time unless there is a major fall within the 12 months.
Projecting Time Horizon Performance Out 12 Months
Based on the various index values over the last 4 years we can project the various time horizon performances out 12 months at a range of 1 year performances to see how they would impact on the various time horizon performances at that time. I have modeled the performance possibilities for the 3rd anniversary of the market bottom. This was inspired by work Doug Short did on future outcomes after high 2 year returns from a bottom for the US S&P 500.
I have heat mapped the results to show which outcomes would be considered possible buy or sell indicators for each performance horizon. You can see that the exceptional 2 year performance as at 6 March 2011 becomes an exceptional 3 year performance on 6 March 2012 even if there is no growth in the market between those 2 dates.
A one year performance of around 10 percent for each of the next few years would not bring most performance horizons into dangerous territory (other than the one representing the flow through of the current exceptional 2 year performance).
Projecting Out 2 Years to the 4th Anniversary of the 2009 Bottom.
At the suggestion of Adam Butler I have projected out 2 years using simple annual compounding of growth. The chart is similar to that above but has rows to show the index value after 1 year and after 2 years both based on an index value of 4896 on 6 March 2011, the second anniversary of the bottom.
Based on this one would not expect growth of more than about 5 to 10% compound (which would keep most horizons performance within the central range of their percentiles. However if one expects that recoveries generally lead to higher percentile performance before a significant market fall, then one might expect performance of around 15% pa resulting in horizons being in the moderately uncommon range (of 65th to 80th percentile for 1, 2, 3 and 4 year rolling returns). The dramatic 1 year recovery from the bottom will be reflected as a very uncommon 4 year performance, while the 5 year performance will still be uncommonly low because it still contains part of the 2008 crash. This is not a prediction, just an analysis of what is likely and unlikely. Watching other indicators of market tops such as yield curve inversion, rising industrial stocks, reduced production, short moving averages crossing below long ones, PE ratios becoming uncommonly high, PE ratios being more than say 1.5 times the inverse of the 10 year bond yield, Shiller's PE 10 exceeding 20, increased fiscal austerity is also necessary to help determine when to rebalance away from equities.
Percentile Performance at Tops
Let's look at each of the major market tops and see what was the percentile performance for each of the 5 time horizons.
Comparing the percentiles performance for each horizon as of 5 April to the percentile performance of prior tops one would conclude that it is highly unlikely that we are at a new top at present as apart from the 2 year performance all time horizons are well below the medians and averages of prior tops. Three items in the chart give some pause for thought.
1. A 61% rise from the bottom is higher than the median and about average
2. The 2 year performance of 38% is up about the 80th percentile
3. At 2.1 years since the bottom, that is at the average period of time to the next top.
4. At 3.4 years since the peak of 2007, the length of this bull market is longer than both average and the median.
Overall, it looks more likely that we have not yet reached a major top.
Percentile Performance at Bottoms
Let's look now at each of the major market bottoms and see what was the percentile performance for each of the 5 time horizons.
Based on the fall since the last top and the 3 and 4 year performance the market still looks like a major bottom as at 5 April. On the other hand, based on the rise since the bottom and the 1 and 2 year performance the market looks like it has recovered away from the bottom.
Buy, Hold or Sell - Time to Rebalance?
Several US commentators and fund managers suggest that the US market is overvalued on many measures including Tobin Q, Shillers PE10, and Corporate Margins (such analysis seems less freely available for the Australian market but please me know if there are free sources for this information).
There are also concerns for the end of QE2, moves to fiscal austerity (a big move in UK and much of the EUR community and those that aspire to membership or wish to borrow from the IMF as a result of banking crises brought on by housing bubbles) and the possibility of rising inflation bringing rising interest rates as in China, India and the EUR community.
With US unemployment still very high at over 8% and the participation rate lower than historical highs, with US housing double dipping in many markets and capacity utilisation well below historical highs and with the yield curve very steep (not inverted as often happens before major crashes), there are many reasons to expect that US fiscal and monetary policy will remain expansionary to support both the real economy and some asset prices like commodities and the stock market (although wealth effect has little positive correlation to borrowing and spending). With Oil over USD110 and challenges to ruling autocrats in oil producing countries and with many commodities at long term highs (largely as a result of the USD at long term lows) some inflation seems likely in the US and consumers will face reduced discretionary consumption as gas/petrol prices rise given the relatively inelastic demand.
The question Richard Koo would ask is whether the US will contract fiscally, reducing private sector income and consumption and, one might expect although the correlation is not always strong, the stock market or will it recognise that in a balance sheet recession private deleveraging is saving which in the absence of changes in the international balance must represent public deficits if GDP is to be maintained.
The answer to Koo would depend largely on US politics. Most "respected" economists and commentators don't recognise that with a fiat currency the US government can adopt the Japanese solution to managing the fall out from the Tech (2000), housing and CRE (commercial real estate) bubbles and the Tea Party and many republicans and even progressives don't understand/accept modern monetary theory and treat the US government as an income and debt constrained "household" (or they don't think it can be explained to and accepted by the public/voters and so are using fear of deficits and debt to gain political advantage/avoid political loss).
While based on many factors the case can be made to remain 100% invested in the market, there are many factors as discussed above which suggest caution.
While it might be largely explained by the size of the fall in 2007-9, one thing which is striking is that when a major bottom occurs the median rise from the prior bottom is only 16% and the largest it has been in the 7 bottoms over 20 years since the 1987 bottom has been 35%, not including the current recovery which has seen a rise of 61% from the 2009 bottom. On this basis it is possibly the time to rebalance a substantial part of your portfolio away from stocks and in favour of short term cash (possibly using bank and credit union "special" term deposit deals) on the basis that even if you miss some proportion of the upside, there is a very good chance you can buy back in at a lower index value and, after deducting interest earned on the proportion switched, the stock market returns, if any, lost by switching part to cash after a recovery of over 2 years have, since the 1987 bottom, been small ( I hope to do more work on this for a future article). You will then have a chance to rebalance in favour of equities near a bottom and ride the next recovery.
The chart of Percentile Performance at Major Bottoms (above) will help you with bottom picking and could be used in conjunction with eg 30 and 90 day moving averages or similar (eg buy when the 30 day rises above the 90 day after a 20% fall.)
A Note on Currency
Remember that Private = Goverment + International for a given level of GDP. This is relevant not only to the need for expansionary fiscal policy during private sector deleveraging (Koo), but also to possible strategies to improve employment and capacity utilisation through currency devaluation, which also leads to changes in USD profits for US multinationals.
Countries with low capacity utilisation could theoretically be expected to export their way out of trouble by lowering interest rates along the curve through QE, causing a weakening currency and stimulating demand for domestic production by making imported goods more expensive.
Clearly not every country can export it's way out of trouble at once. Because of pegging to the US dollar by many countries including China, the US might be considered very unlikely to be able to export its way out of trouble. Australia is suffering from the Dutch Disease. It's high currency resulting largely from high commodity prices, comparatively high short and medium term interst rates and its low level of government (but not private) debt is rendering it's import competing businesses outside of mining much less competitive and reducing the value in AUD of profits earnt offshore. Conversely the US profits of multi nationals are being helped by the fall in the value of the USD against some currencies.
More sophisticated investors might like to consider returns of various markets benchmarked against a single currency. For example, the return to AUD investors has had two components. The currency strengthening to USD105 and the rise in the All Ords. The US investor has, in EUR and Dollar Index terms, had part of the US stock market performance offset by losses in the USD.
For Australian investors there is some uncertainty about the capacity of China (and India) to maintain it's growth without inflation which could lead to falls in commodity prices unless any slowing in domestic investment and consumption is offset by growth in the export sector as a result of eg US recovery extending to "Main Street".
Any international fall in markets would likely be accompanied by a fall in production, commodity demand and prices and the AUD. In March 2009 the All Ords was down almost 80% in USD terms compared to over 50% in AUD terms. Australian investors rebalancing from the All Ords could consider foreign currency denominated assets including short term deposits, but this exposes the investor to currency losses as well as gains. If a clear likely top is identified a la 2007, (long bull run, all time horizon performances in high percentiles) then a switch to eg USD or German medium to long term bonds could be very rewarding. See my earlier article: Currency and Diversification Impacts on Investment Returns to Australian Investors for full analysis.
ETF's (Exchange Traded Funds) based on regions, sectors or commodities can be used to diversify on an unhedged basis using either the US or Australian markets. Care should be taken to understand the impacts of different tax laws including the benefits to Australian investors of dividend imputation when invested in Australian stocks, and superannuation laws and taxation where relevant.
This blog is for informational and educational purposes only and does not represent investment advice.
Some sources of information and hat tips for commentary and analysis or leads to same include:
John Hussman - http://www.hussmanfunds.com/weeklyMarketComment.html
Adam Butler - http://www.butlerphilbrick.com/CaseStudies.html
Doug Short - http://www.dshort.com/
Calculated Risk - http://www.calculatedriskblog.com/
Paper Economy - http://paper-money.blogspot.com/
Pragmatic Capitalist - http://pragcap.com/
Bill Mitchell - http://bilbo.economicoutlook.net/blog/
Steve Keen - http://www.debtdeflation.com/blogs/
Business insider - http://www.businessinsider.com/
Yahoo Finance provides data for the Australian All Ordinaries Index since 1984.
I have used that data to analyse the Australian Market into percentiles of performance. I have then looked at those percentiles at peaks and troughs. I have also looked at what the percentiles of 1, 2, 3, 4 and 5 year growth would be at the 3rd anniversary of the market bottom on 6 March 2009 at different levels of performance over the next 12 months to see what the market would then look like against historical peaks and troughs. I have started some of the analysis since 1984, but some has been constrained because the time performance can oly be calculated at the relevant number of years after the first index value is available.
I have also charted each recovery since from the 1987 bottom in percentage terms to be able to compare the size and speed of recovery in percentage terms by number of trading days since the bottom.
My goal is to provide useful guidance as to the likelihood of future performance, based on the historical recoveries and performance of the Australian All Ordinaries since 1984.
Comparative Speed and Scale of Recovery from March 2009
After 531 trading days, every other recovery since 1984 was lower in percentage terms than today's recovery of 62.1% from the 2009 low. The only market since 1984 which had a higher percentage recovery at any time within 531 days was the recovery from 1992 which had recovered 71.8% by trading day 310, but after 531 days had fallen back to a recovery of only 39.8%.
Looking at it another way, the current recovery is 70 days ahead of the 2003 recovery (the next strongest) and 319 days ahead of the average recovery since the 1987 bottom.
Looking out to day 800 (chosen as it is the point with the widest range of percentage recoveries since 1984 and only170 trading days away from now for this recovery):
1. the 2003 recovery was at 97.4% (best) (an implied future rise of 36.3 points to an index value of 7993 from today's intraday value of 5044, an index rise of 58.4%)
2. the 1987 recovery was at 5.9% (worst) (an implied future fall of 56.2 points to an index value of 3420, an index fall of 32.2%
3 the average of all recoveries was at 60.1% (implying an almost insignificant fall from today).
Of the 5 other recoveries from 20% falls since the 1987 bottom, 3 had ended and had 20% falls by day 531. Only the recoveries from the 1995 and 2003 bottoms went for longer without a 20% fall. The biggest fall in this recovery has been the 11.3% fall of 567.4 points from 5024.1 on 15 April to 4456.7 on 15 July 2010.
Historical Percentile Performance of the Australian Market
The chart below shows the percentiles of performance over rolling 1 to 5 year periods since 1991. It was compiled based on performance to 5 April 2010. From the chart it can be seen that the 5 year performance is between -10% and 50% about 54% of the time. Similarly it can be seen that 1 year performance is less than 0 about 30% of the time.
From this you might assume that at any point in time the likelihood is for positive growth over each of these time horizons, but this ignores that the strongest growth, at least for 1 and 2 year horizons, generally follows a market bottom eg after 2009. From the chart though it seems that if all horizons are in the say lowest 10 percentiles (1 to 10) it is likely a market bottom and if they are all in the highest say 10 percentiles (90 to 100) it is likely a market top. The starting date used (1991) is to ensure all time horizons are looked at from the same starting date, in this case near the bottom of the post 1987 performance for all 5 horizons.
Note that the longer the horizon the more the performance can lag major changes in the index. Note that at present while the All Ords are up 61% from the 2009 bottom it is only reflected in the green 2 year performance with all other horizons showing relatively poor performance.
Percentile Heat Map
We can get a general impression of what are good and bad levels of performance over different time horizons by looking at a percentile heat map. If most performance horizons are in the top 10 percentiles it's probably time to consider rebalancing to cash or bonds, or lengthen duration of bonds, particularly if the yield curve is inverted. On the other hand, if the yield curve is unusually steep and most performance horizons are in the lowest decile then it is time to consider rebalancing toward equities and perhaps, if you have modeled the range of possible negative outcomes and have recent experience, even considering buying call options, geared equities or contracts for difference.
The heat map below tells you that since 1991 a 1 year performance of -14.6% or less is in the bottom 10% of outcomes for 1 year performance and is a likely buying opportunity. Similarly a 5 year performance of 81% or better is better than 90% of results and it is likely a dangerous time to buy or hold.
Recognise however that because of lags in the longer time horizons a 3, 4 or 5 year performance can still be quite good even as the market approaches a bottom and a good 2 year performance now will likely be translated into a good 3 year performance in 12 months time unless there is a major fall within the 12 months.
Projecting Time Horizon Performance Out 12 Months
Based on the various index values over the last 4 years we can project the various time horizon performances out 12 months at a range of 1 year performances to see how they would impact on the various time horizon performances at that time. I have modeled the performance possibilities for the 3rd anniversary of the market bottom. This was inspired by work Doug Short did on future outcomes after high 2 year returns from a bottom for the US S&P 500.
I have heat mapped the results to show which outcomes would be considered possible buy or sell indicators for each performance horizon. You can see that the exceptional 2 year performance as at 6 March 2011 becomes an exceptional 3 year performance on 6 March 2012 even if there is no growth in the market between those 2 dates.
A one year performance of around 10 percent for each of the next few years would not bring most performance horizons into dangerous territory (other than the one representing the flow through of the current exceptional 2 year performance).
Projecting Out 2 Years to the 4th Anniversary of the 2009 Bottom.
At the suggestion of Adam Butler I have projected out 2 years using simple annual compounding of growth. The chart is similar to that above but has rows to show the index value after 1 year and after 2 years both based on an index value of 4896 on 6 March 2011, the second anniversary of the bottom.
Based on this one would not expect growth of more than about 5 to 10% compound (which would keep most horizons performance within the central range of their percentiles. However if one expects that recoveries generally lead to higher percentile performance before a significant market fall, then one might expect performance of around 15% pa resulting in horizons being in the moderately uncommon range (of 65th to 80th percentile for 1, 2, 3 and 4 year rolling returns). The dramatic 1 year recovery from the bottom will be reflected as a very uncommon 4 year performance, while the 5 year performance will still be uncommonly low because it still contains part of the 2008 crash. This is not a prediction, just an analysis of what is likely and unlikely. Watching other indicators of market tops such as yield curve inversion, rising industrial stocks, reduced production, short moving averages crossing below long ones, PE ratios becoming uncommonly high, PE ratios being more than say 1.5 times the inverse of the 10 year bond yield, Shiller's PE 10 exceeding 20, increased fiscal austerity is also necessary to help determine when to rebalance away from equities.
Percentile Performance at Tops
Let's look at each of the major market tops and see what was the percentile performance for each of the 5 time horizons.
1. A 61% rise from the bottom is higher than the median and about average
2. The 2 year performance of 38% is up about the 80th percentile
3. At 2.1 years since the bottom, that is at the average period of time to the next top.
4. At 3.4 years since the peak of 2007, the length of this bull market is longer than both average and the median.
Overall, it looks more likely that we have not yet reached a major top.
Percentile Performance at Bottoms
Let's look now at each of the major market bottoms and see what was the percentile performance for each of the 5 time horizons.
Based on the fall since the last top and the 3 and 4 year performance the market still looks like a major bottom as at 5 April. On the other hand, based on the rise since the bottom and the 1 and 2 year performance the market looks like it has recovered away from the bottom.
Buy, Hold or Sell - Time to Rebalance?
Several US commentators and fund managers suggest that the US market is overvalued on many measures including Tobin Q, Shillers PE10, and Corporate Margins (such analysis seems less freely available for the Australian market but please me know if there are free sources for this information).
There are also concerns for the end of QE2, moves to fiscal austerity (a big move in UK and much of the EUR community and those that aspire to membership or wish to borrow from the IMF as a result of banking crises brought on by housing bubbles) and the possibility of rising inflation bringing rising interest rates as in China, India and the EUR community.
With US unemployment still very high at over 8% and the participation rate lower than historical highs, with US housing double dipping in many markets and capacity utilisation well below historical highs and with the yield curve very steep (not inverted as often happens before major crashes), there are many reasons to expect that US fiscal and monetary policy will remain expansionary to support both the real economy and some asset prices like commodities and the stock market (although wealth effect has little positive correlation to borrowing and spending). With Oil over USD110 and challenges to ruling autocrats in oil producing countries and with many commodities at long term highs (largely as a result of the USD at long term lows) some inflation seems likely in the US and consumers will face reduced discretionary consumption as gas/petrol prices rise given the relatively inelastic demand.
The question Richard Koo would ask is whether the US will contract fiscally, reducing private sector income and consumption and, one might expect although the correlation is not always strong, the stock market or will it recognise that in a balance sheet recession private deleveraging is saving which in the absence of changes in the international balance must represent public deficits if GDP is to be maintained.
The answer to Koo would depend largely on US politics. Most "respected" economists and commentators don't recognise that with a fiat currency the US government can adopt the Japanese solution to managing the fall out from the Tech (2000), housing and CRE (commercial real estate) bubbles and the Tea Party and many republicans and even progressives don't understand/accept modern monetary theory and treat the US government as an income and debt constrained "household" (or they don't think it can be explained to and accepted by the public/voters and so are using fear of deficits and debt to gain political advantage/avoid political loss).
While based on many factors the case can be made to remain 100% invested in the market, there are many factors as discussed above which suggest caution.
While it might be largely explained by the size of the fall in 2007-9, one thing which is striking is that when a major bottom occurs the median rise from the prior bottom is only 16% and the largest it has been in the 7 bottoms over 20 years since the 1987 bottom has been 35%, not including the current recovery which has seen a rise of 61% from the 2009 bottom. On this basis it is possibly the time to rebalance a substantial part of your portfolio away from stocks and in favour of short term cash (possibly using bank and credit union "special" term deposit deals) on the basis that even if you miss some proportion of the upside, there is a very good chance you can buy back in at a lower index value and, after deducting interest earned on the proportion switched, the stock market returns, if any, lost by switching part to cash after a recovery of over 2 years have, since the 1987 bottom, been small ( I hope to do more work on this for a future article). You will then have a chance to rebalance in favour of equities near a bottom and ride the next recovery.
The chart of Percentile Performance at Major Bottoms (above) will help you with bottom picking and could be used in conjunction with eg 30 and 90 day moving averages or similar (eg buy when the 30 day rises above the 90 day after a 20% fall.)
A Note on Currency
Remember that Private = Goverment + International for a given level of GDP. This is relevant not only to the need for expansionary fiscal policy during private sector deleveraging (Koo), but also to possible strategies to improve employment and capacity utilisation through currency devaluation, which also leads to changes in USD profits for US multinationals.
Countries with low capacity utilisation could theoretically be expected to export their way out of trouble by lowering interest rates along the curve through QE, causing a weakening currency and stimulating demand for domestic production by making imported goods more expensive.
Clearly not every country can export it's way out of trouble at once. Because of pegging to the US dollar by many countries including China, the US might be considered very unlikely to be able to export its way out of trouble. Australia is suffering from the Dutch Disease. It's high currency resulting largely from high commodity prices, comparatively high short and medium term interst rates and its low level of government (but not private) debt is rendering it's import competing businesses outside of mining much less competitive and reducing the value in AUD of profits earnt offshore. Conversely the US profits of multi nationals are being helped by the fall in the value of the USD against some currencies.
More sophisticated investors might like to consider returns of various markets benchmarked against a single currency. For example, the return to AUD investors has had two components. The currency strengthening to USD105 and the rise in the All Ords. The US investor has, in EUR and Dollar Index terms, had part of the US stock market performance offset by losses in the USD.
For Australian investors there is some uncertainty about the capacity of China (and India) to maintain it's growth without inflation which could lead to falls in commodity prices unless any slowing in domestic investment and consumption is offset by growth in the export sector as a result of eg US recovery extending to "Main Street".
Any international fall in markets would likely be accompanied by a fall in production, commodity demand and prices and the AUD. In March 2009 the All Ords was down almost 80% in USD terms compared to over 50% in AUD terms. Australian investors rebalancing from the All Ords could consider foreign currency denominated assets including short term deposits, but this exposes the investor to currency losses as well as gains. If a clear likely top is identified a la 2007, (long bull run, all time horizon performances in high percentiles) then a switch to eg USD or German medium to long term bonds could be very rewarding. See my earlier article: Currency and Diversification Impacts on Investment Returns to Australian Investors for full analysis.
ETF's (Exchange Traded Funds) based on regions, sectors or commodities can be used to diversify on an unhedged basis using either the US or Australian markets. Care should be taken to understand the impacts of different tax laws including the benefits to Australian investors of dividend imputation when invested in Australian stocks, and superannuation laws and taxation where relevant.
This blog is for informational and educational purposes only and does not represent investment advice.
Some sources of information and hat tips for commentary and analysis or leads to same include:
John Hussman - http://www.hussmanfunds.com/weeklyMarketComment.html
Adam Butler - http://www.butlerphilbrick.com/CaseStudies.html
Doug Short - http://www.dshort.com/
Calculated Risk - http://www.calculatedriskblog.com/
Paper Economy - http://paper-money.blogspot.com/
Pragmatic Capitalist - http://pragcap.com/
Bill Mitchell - http://bilbo.economicoutlook.net/blog/
Steve Keen - http://www.debtdeflation.com/blogs/
Business insider - http://www.businessinsider.com/
Tuesday, June 8, 2010
First, no credit card debt and second, pay off non-deductible debt!
Generally the best returns are to pay off all non-deductible debt. First, credit cards, second, personal loans and third, home mortgage.
Better still, avoid credit card debt like the plague!!
Also, always save a hefty deposit on anything you might otherwise finance on a personal loan and buy what you need, not what you might like to have - saving $5000 before buying and buying a $5,000 cheaper car so you borrow $10,000 less can give you a substantial interest saving, particularly if you borrow over a shorter period (making higher repayments) as well.
Assume you have a debt of $1000.
To pay $200 of 20% non-deductible interest on a credit card debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you have to earn 333.33 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 200/(100-40)*100 = 333.33
To pay $150 of 15% non-deductible interest on a personal loan of $1000 (eg for a car) out of after tax earnings if you are on a 40% marginal tax rate you have to earn 250.00 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 150/(100-40)*100 = 250.00
To pay $100 of 10% non-deductible interest on a mortgage debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you have to earn 166.67 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 100/(100-40)*100 = 166.67
To pay $100 of 10% tax deductible interest on a mortgage debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you only have to earn $100.00 because you get a tax deduction and so don’t have to earn extra to pay the tax on the extra earnings.
In summary, the amount you have to earn to pay one year's interest on a $1,000 loan using the examples above is :
1. Credit card 20%: 333
2. Personal loan 15%: 250
3 Home loan 10%: 167
4. Investment loan 10%: 100
Another way of looking at this is to gross up the interest rate to a pre-tax rate:
1. credit card rate is 33.3%
2. personal loan rate is 25%
3. home loan rate is 16.7%
4. investment loan rate is 10%
So generally the greatest after tax return comes from paying off non-deductible debt, and it is worth ensuring that you have sufficient spare cash, even in very low interest paying accounts, that you never have to pay interest on your credit card.
This analysis can be complicated by taking capital gains into account on investments and depending on anticipated capital gain it might be worth using your surplus to invest in stocks rather than pay off your non-deductible home loan if the market has just fallen 50% and it has turned up and just broken through the 100 day SMA.
It may even be worth borrowing more against any “surplus” equity you have in your home to make such an investment, even though you are not paying down non-deductible debt as fast as you could.
But to keep it simple for now, first pay off all non-deductible debt, starting with your credit card, then personal loans, then home loans, and when you only have your home loan left, use an offset account so your savings are all offsetting your home loan until you need them to pay bills.
If you need to borrow using a personal loan eg for a car, make sure that your lender will allow you to pay it off faster without any additional costs. This may mean that you need to ensure that the loan is a floating rate loan. This way you can take a longer loan period to give yourself a margin of safety, but pay it off fast.
The golden rules are:
Credit Cards
No credit card debt
a) never spend on a credit card anything you can't pay off in the interest free period
b) never take a cash advance from a credit card,
c) in fact run your with a small credit balance
Debt and Investing
Pay off all non-deductible debt first
Better still, avoid credit card debt like the plague!!
Also, always save a hefty deposit on anything you might otherwise finance on a personal loan and buy what you need, not what you might like to have - saving $5000 before buying and buying a $5,000 cheaper car so you borrow $10,000 less can give you a substantial interest saving, particularly if you borrow over a shorter period (making higher repayments) as well.
Assume you have a debt of $1000.
To pay $200 of 20% non-deductible interest on a credit card debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you have to earn 333.33 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 200/(100-40)*100 = 333.33
To pay $150 of 15% non-deductible interest on a personal loan of $1000 (eg for a car) out of after tax earnings if you are on a 40% marginal tax rate you have to earn 250.00 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 150/(100-40)*100 = 250.00
To pay $100 of 10% non-deductible interest on a mortgage debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you have to earn 166.67 (Amount to be paid divided by (100 – marginal tax rate) multiplied by 100: 100/(100-40)*100 = 166.67
To pay $100 of 10% tax deductible interest on a mortgage debt of $1000 out of after tax earnings if you are on a 40% marginal tax rate you only have to earn $100.00 because you get a tax deduction and so don’t have to earn extra to pay the tax on the extra earnings.
In summary, the amount you have to earn to pay one year's interest on a $1,000 loan using the examples above is :
1. Credit card 20%: 333
2. Personal loan 15%: 250
3 Home loan 10%: 167
4. Investment loan 10%: 100
Another way of looking at this is to gross up the interest rate to a pre-tax rate:
1. credit card rate is 33.3%
2. personal loan rate is 25%
3. home loan rate is 16.7%
4. investment loan rate is 10%
So generally the greatest after tax return comes from paying off non-deductible debt, and it is worth ensuring that you have sufficient spare cash, even in very low interest paying accounts, that you never have to pay interest on your credit card.
This analysis can be complicated by taking capital gains into account on investments and depending on anticipated capital gain it might be worth using your surplus to invest in stocks rather than pay off your non-deductible home loan if the market has just fallen 50% and it has turned up and just broken through the 100 day SMA.
It may even be worth borrowing more against any “surplus” equity you have in your home to make such an investment, even though you are not paying down non-deductible debt as fast as you could.
But to keep it simple for now, first pay off all non-deductible debt, starting with your credit card, then personal loans, then home loans, and when you only have your home loan left, use an offset account so your savings are all offsetting your home loan until you need them to pay bills.
If you need to borrow using a personal loan eg for a car, make sure that your lender will allow you to pay it off faster without any additional costs. This may mean that you need to ensure that the loan is a floating rate loan. This way you can take a longer loan period to give yourself a margin of safety, but pay it off fast.
The golden rules are:
Credit Cards
No credit card debt
a) never spend on a credit card anything you can't pay off in the interest free period
b) never take a cash advance from a credit card,
c) in fact run your with a small credit balance
Debt and Investing
Pay off all non-deductible debt first
Sunday, June 6, 2010
When do I buy back in?
At the moment I am only about 30% invested in the stock market.
The question that is exercising my mind is "how will I know when to buy back in?
I am looking at using a cross of the 30 day Simple Moving Average (SMA) by the 10 day SMA from below (subject to the price being higher than the cross). The market has recently fallen up to 14% and is currently still down 10%. Unless we have a dramatic fall on Monday in response to the 3.4% fall on Wall Street S&P500 last Friday, the cross could happen this coming week.
There are risks to this approach. The risk is that the signals reverse several times without the market moving much between the crosses. That is called being whipsawn. It can cause losses and each transaction also has costs. The market is described as "trending sideways""or "in transition".
However, you need something to help you judge when to buy back in, or increase your weighting of stocks. Some would say the 200 SMA is more reliable, but you have much bigger losses in a major downturn or lost opportunities in a major upturn using such a long average.
So why would I use the 10 and 30 cross? Because it can be quite a profitable indicator. (If the market was down 25% I would likely use a much longer SMA than 30, maybe 100 - I feel that the deeper the fall, the longer the SMA for a buy signal should be to avoid whipsawing, but I have no backtesting to support this feel):
It gave a 10% rise from October 2001 to February 2002 with no whipsaws, and a 17% rise from March 2003 to March 2004 with 5 whipsaws. Using a 50 day SMA instead of a 30 day SMA would have given a 16% rise with no whipsaws and would have been a better choice on this example as the rise was gradual with 5 pullbacks. A sudden rise with no pullbacks would have been well suited to a shorter SMA like 30 Trading Days.
The chart above also shows how you can get whipsawn as the market goes through a 20% decline. If you used a 120 SMA instead of 30 you would have been whipsawn once on the way down and not at all on the way up. you would have been down 3% before your first sell signal, have lost 2% in the whipsaw and then picked up 14% for a net 9% gain compared to a zero gain over the same period if you had just continued to hold your stocks all through the period.
Here is a chart with two significant rises where there was no whipsawing using the 10 and 30 SMA cross. Note that we are down a similar amount now as was the case from both of the bottoms shown in this chart.
But what of the risks of multiple whipsaws with losses and transaction costs? Right click and open the next chart in a new window.
From Jan 91 to July 92 you would have had 7 roundtrip transactions for a total market movement of only +22%. a number of the round trips would have been loss making, reducing the 22% gain to one where on a net basis you captured only between 1/2 and 3/4 of the 22%. A 200 day SMA instead of a 30 would have given less whip sawing, but a higher first buy price, so there may not have been much benefit in using a longer SMA.
So not buying would have led to missing a profit of between 10 and 15 % in 18 months compared to being whipsawn by the 10 and 30 sma while a perfect buy at the bottom would have given 22% return - but how would you pcik the absolute bottom?
No one rings a bell at the top or bottom of the market! You have to have some basis on which to buy and sell, or adjust your portfolio weightings.
Choose the lengths of the SMA's you will use based on your own costs of transaction, loss aversion, ability to replenish funds and consider other factors as well, including fundamentals.
Before embarking on any strategy involving moving averages you should have a look at some recent work by Doug Short::
http://dshort.com/articles/2010/Nikkei-monthly-moving-averages.html
http://dshort.com/articles/2010/SP500-monthly-moving-average-history.html
The question that is exercising my mind is "how will I know when to buy back in?
I am looking at using a cross of the 30 day Simple Moving Average (SMA) by the 10 day SMA from below (subject to the price being higher than the cross). The market has recently fallen up to 14% and is currently still down 10%. Unless we have a dramatic fall on Monday in response to the 3.4% fall on Wall Street S&P500 last Friday, the cross could happen this coming week.
There are risks to this approach. The risk is that the signals reverse several times without the market moving much between the crosses. That is called being whipsawn. It can cause losses and each transaction also has costs. The market is described as "trending sideways""or "in transition".However, you need something to help you judge when to buy back in, or increase your weighting of stocks. Some would say the 200 SMA is more reliable, but you have much bigger losses in a major downturn or lost opportunities in a major upturn using such a long average.
So why would I use the 10 and 30 cross? Because it can be quite a profitable indicator. (If the market was down 25% I would likely use a much longer SMA than 30, maybe 100 - I feel that the deeper the fall, the longer the SMA for a buy signal should be to avoid whipsawing, but I have no backtesting to support this feel):
It gave a 10% rise from October 2001 to February 2002 with no whipsaws, and a 17% rise from March 2003 to March 2004 with 5 whipsaws. Using a 50 day SMA instead of a 30 day SMA would have given a 16% rise with no whipsaws and would have been a better choice on this example as the rise was gradual with 5 pullbacks. A sudden rise with no pullbacks would have been well suited to a shorter SMA like 30 Trading Days.The chart above also shows how you can get whipsawn as the market goes through a 20% decline. If you used a 120 SMA instead of 30 you would have been whipsawn once on the way down and not at all on the way up. you would have been down 3% before your first sell signal, have lost 2% in the whipsaw and then picked up 14% for a net 9% gain compared to a zero gain over the same period if you had just continued to hold your stocks all through the period.
Here is a chart with two significant rises where there was no whipsawing using the 10 and 30 SMA cross. Note that we are down a similar amount now as was the case from both of the bottoms shown in this chart.
But what of the risks of multiple whipsaws with losses and transaction costs? Right click and open the next chart in a new window.
From Jan 91 to July 92 you would have had 7 roundtrip transactions for a total market movement of only +22%. a number of the round trips would have been loss making, reducing the 22% gain to one where on a net basis you captured only between 1/2 and 3/4 of the 22%. A 200 day SMA instead of a 30 would have given less whip sawing, but a higher first buy price, so there may not have been much benefit in using a longer SMA.
So not buying would have led to missing a profit of between 10 and 15 % in 18 months compared to being whipsawn by the 10 and 30 sma while a perfect buy at the bottom would have given 22% return - but how would you pcik the absolute bottom?
No one rings a bell at the top or bottom of the market! You have to have some basis on which to buy and sell, or adjust your portfolio weightings.
Choose the lengths of the SMA's you will use based on your own costs of transaction, loss aversion, ability to replenish funds and consider other factors as well, including fundamentals.
Before embarking on any strategy involving moving averages you should have a look at some recent work by Doug Short::
http://dshort.com/articles/2010/Nikkei-monthly-moving-averages.html
http://dshort.com/articles/2010/SP500-monthly-moving-average-history.html
Thursday, June 3, 2010
Currency and Diversification Impacts on Investment Returns to Australian Investors
I have used the MSCI Barra performance indices (http://www.mscibarra.com/products/indices/international_equity_indices/gimi/stdindex/performance.html) for some of this article as it allows comparisons between performance measured in EUR, USD and local currency. Some charts are from the free version of Incredible Charts (http://incrediblecharts.com).
The Australian Market in AUD and USD
We know what happened with the Australian All Ordinaries from 2007 to 1 June 2010.
At the bottom of the Australian Stock Market in March 2009 the MSCI Barra index was off 53% in AUD terms from 30 October 2007.
In USD terms it was off about 77%!
The recovery in the MSCI Barra index for Australia from the 9 March 2009 bottom to 12 April 2010 was 58%!
In USD terms it was up 132%
These differences are because the AUD/USD Exchange rate was also changing over the same period. Chart courtesy of Incredible Charts free charting software.
So now we can look at the Australian Stock Market in each of AUD and USD using MSCI Barra.
The numbers above illustrate the difference that movements in the currency can have on investment returns.
Currency and Switching Asset Classes
I am now going to use the perfect example to make a point, but nobody could possibly have achieved the result so don’t think that the actual return is even remotely possible. (I will ignore dividends and interest – at 4% pa they have little effect over the time period we are considering and partially offset one another, even after tax and compounding effect.)
An Australian investor who switched from the All Ords to USD cash deposit on 30 October and then switched back to the All Ords on 9 March 2009 and then switched back to USD cash on 14 April would have :
At end October 2007 for AUD100 he would have got USD 93
At 9 March 2009 for USD93 he would have got AUD 1.56 for each USD or AUD145
As at April 14 2010 that AUD145 in the Australian Stock market since 9 March 2009 would be worth AUD229
On 14 April that AUD 229 would buy USD212 at 0.93
Today the USD212 is worth AUD256
An investor with AUD 100 in the All Ords on 30 October 2007 who has just held his investment now has AUD65
The perfect switching strategy to USD cash and back twice would give you 3.9 times as much AUD now as a mere buy and hold strategy! (But remember I said this was totally theoretical and unachievable in practice.)
There could have been additional profit if our hypothetical switching investor had moved to US 10 year bonds instead of cash. As the interest rates on US 10 year Treasury notes went lower during 2008/09 (see chart of $TNX below from StockCharts.com) the price of 10 year bonds bought in October 2007 would have increased. As rates increased in 2009/10 the price would have fallen back but as rates are still lower the price would still be higher than October 2007.
Of course it works the other way too. A US investor who held his Australian All Ords at US93 (AUD100 equivalent) on 30 October 2007, sold at the bottom on 9 March 2009 and took his funds back into USD and switched to cash would have only USD21 (AUD23 ) on 14 April 2009, compared to the Australian investor who did nothing with AUD65 – about 3 times as much as the US investor.
So the point is, if the AUD is has had large price increases compared to most currencies, maybe switching, or increasing your exposure, to those other currencies for part of your investment might make sense (and vice versa).
Similarly, if the stock market (or any other asset class) has grown very quickly for a long time, switching to, or increasing your exposure to, non-correlated assets that are at lower prices might save a significant fall in investment value if the one which has risen starts to fall.
Diversification to Non-Correlated Assets
Now imagine if our Australian Investor had 50% in USD cash and 50% in AUD shares.
At the peak he had AUD50 in shares and AUD50 (USD46.5) in cash. At the bottom of the AUD market he would have AUD23 in AUD Shares and AUD72.54 in USD Cash of USD46.5. His AUD100 is down to only AUD95.5 instead of being down to AUD47 as it would be in shares. The diversification into non-correlated asset classes has dramatically reduced the unrealised loss as at 9 March 2010.
At 14 April the position is AUD 32.5 in Shares and the USD46.5 is worth AUD50 again (by absolute coincidence of the exchange rates at 1 AUD = 0.93USD on both 30 October 2007 and 14 April 2010). Total value AUD 82.5. On this occasion the diversified investor is in front of the AUD shares only investor, but the generally expected result is that the diversified investor will have sacrificed some earnings but will have reduced volatility as shown above by the much lower fall in value of investment.
Conclusion
This is not a suggestion to trade, but is an introduction to three concepts:
1. Foreign currency exposure can be a good thing.
2. Considering rebalancing strategic asset allocation from time to time might be worthwhile. (called Active or Tactical Asset Allocation – these terms mean different things to different people. I am talking about 4 times a year style approach, not necessarily exactly quarterly, maybe at times of crisis in some country or region, or after a fall in stock markets in a region of more than say, 8%.)
3. Having a portfolio of non correlated assets (like USD bonds and AUD stocks) can reduce the volatility in the value of your investment and help you sleep nights.
Next Post
My next post will be on portfolio construction, risk aversion, volatility and diversification.
The Australian Market in AUD and USD
We know what happened with the Australian All Ordinaries from 2007 to 1 June 2010.
At the bottom of the Australian Stock Market in March 2009 the MSCI Barra index was off 53% in AUD terms from 30 October 2007.
In USD terms it was off about 77%!
The recovery in the MSCI Barra index for Australia from the 9 March 2009 bottom to 12 April 2010 was 58%!
In USD terms it was up 132%
These differences are because the AUD/USD Exchange rate was also changing over the same period. Chart courtesy of Incredible Charts free charting software.
So now we can look at the Australian Stock Market in each of AUD and USD using MSCI Barra.
The numbers above illustrate the difference that movements in the currency can have on investment returns.
Currency and Switching Asset Classes
I am now going to use the perfect example to make a point, but nobody could possibly have achieved the result so don’t think that the actual return is even remotely possible. (I will ignore dividends and interest – at 4% pa they have little effect over the time period we are considering and partially offset one another, even after tax and compounding effect.)
An Australian investor who switched from the All Ords to USD cash deposit on 30 October and then switched back to the All Ords on 9 March 2009 and then switched back to USD cash on 14 April would have :
At end October 2007 for AUD100 he would have got USD 93
At 9 March 2009 for USD93 he would have got AUD 1.56 for each USD or AUD145
As at April 14 2010 that AUD145 in the Australian Stock market since 9 March 2009 would be worth AUD229
On 14 April that AUD 229 would buy USD212 at 0.93
Today the USD212 is worth AUD256
An investor with AUD 100 in the All Ords on 30 October 2007 who has just held his investment now has AUD65
The perfect switching strategy to USD cash and back twice would give you 3.9 times as much AUD now as a mere buy and hold strategy! (But remember I said this was totally theoretical and unachievable in practice.)
There could have been additional profit if our hypothetical switching investor had moved to US 10 year bonds instead of cash. As the interest rates on US 10 year Treasury notes went lower during 2008/09 (see chart of $TNX below from StockCharts.com) the price of 10 year bonds bought in October 2007 would have increased. As rates increased in 2009/10 the price would have fallen back but as rates are still lower the price would still be higher than October 2007.
Of course it works the other way too. A US investor who held his Australian All Ords at US93 (AUD100 equivalent) on 30 October 2007, sold at the bottom on 9 March 2009 and took his funds back into USD and switched to cash would have only USD21 (AUD23 ) on 14 April 2009, compared to the Australian investor who did nothing with AUD65 – about 3 times as much as the US investor.
So the point is, if the AUD is has had large price increases compared to most currencies, maybe switching, or increasing your exposure, to those other currencies for part of your investment might make sense (and vice versa).
Similarly, if the stock market (or any other asset class) has grown very quickly for a long time, switching to, or increasing your exposure to, non-correlated assets that are at lower prices might save a significant fall in investment value if the one which has risen starts to fall.
Diversification to Non-Correlated Assets
Now imagine if our Australian Investor had 50% in USD cash and 50% in AUD shares.
At the peak he had AUD50 in shares and AUD50 (USD46.5) in cash. At the bottom of the AUD market he would have AUD23 in AUD Shares and AUD72.54 in USD Cash of USD46.5. His AUD100 is down to only AUD95.5 instead of being down to AUD47 as it would be in shares. The diversification into non-correlated asset classes has dramatically reduced the unrealised loss as at 9 March 2010.
At 14 April the position is AUD 32.5 in Shares and the USD46.5 is worth AUD50 again (by absolute coincidence of the exchange rates at 1 AUD = 0.93USD on both 30 October 2007 and 14 April 2010). Total value AUD 82.5. On this occasion the diversified investor is in front of the AUD shares only investor, but the generally expected result is that the diversified investor will have sacrificed some earnings but will have reduced volatility as shown above by the much lower fall in value of investment.
Conclusion
This is not a suggestion to trade, but is an introduction to three concepts:
1. Foreign currency exposure can be a good thing.
2. Considering rebalancing strategic asset allocation from time to time might be worthwhile. (called Active or Tactical Asset Allocation – these terms mean different things to different people. I am talking about 4 times a year style approach, not necessarily exactly quarterly, maybe at times of crisis in some country or region, or after a fall in stock markets in a region of more than say, 8%.)
3. Having a portfolio of non correlated assets (like USD bonds and AUD stocks) can reduce the volatility in the value of your investment and help you sleep nights.
Next Post
My next post will be on portfolio construction, risk aversion, volatility and diversification.
Saturday, May 29, 2010
Into and out of Australian Bear Markets since 1984
Using data from Yahoo Finance I have constructed a series of charts which show the lead into the various bear markets in the Australian All Ordinaries and the recovery .
I have also constructed an average of the 5 major bear markets prior to the 2007 bear. This average has then been incorporated in each chart in an arbitrary manner to most closely, in my view, approximate the relevant bear market. The timing of the bottoms has been aligned, but not the actual lowest points. There is also a small amount of vertical compression or expansion in most charts. A green linear trend line for the average recovery is also shown.
This approach makes it easy to see an estimate of the dramatic out or underperformance during different periods. For example, look at the dramatic peaking of the market in 1987 compared to the general trend of the average market leading into a peak and trough.
All charts have the peak before the trough scaled to 1 (100%) so the amount of fall to the bottom can be read from the scale.
Please be aware that in some instances the end of the recovery in one chart may overlap with the period before the trough on another chart eg 1991 trough overlaps on 1992 chart and vice versa.
Click the chart images for a larger image. (Use right click > open in new tab.)
1987
Dramatic bull rally to peak was clearly unsustainable.
1991
The second trough is 1992.
1992
First trough is 1991. 1992 trough is at centre. Note excessively fast recovery to new highs was not sustainable.
1995
Very closely approximates the Average peak, trough and recovery.
2003
The start of the unsustainable rally towards the 2007 peak can be seen towards the end of the recovery from the 2003 trough.
2009
The scale of the fall from 2007 to 2009 makes it very arbitrary as to where the Average recovery is shown. I have chosen to align to the recovery side of the chart. I have also added the graph of the recovery from 1991, also adjusted to align to the recovery side of the chart. Last date is 1 June 2010, day 312, and the correction at the bottom was minus 14.9%. There has been a small rebound since to but is now only minus 11.85%.. The fall may well resume as most countries have broken below their 200 day simple moving average which in some markets is about 65% reliable as an indicator of a major change in trend. This possibility is supported by such fundamentals as:
* a possible second wave of real estate defaults in the US,
* the EUR crisis of sovereign debt owed by the PIIGS (Portugal, Ireland, Italy, Spain)
* the austerity being imposed on those countries
* the continuing possibility that default and withdrawal from the Euro by those countries will be more politically acceptable to voters
* the apparent end of a possible bubble and general slowing of the economy in China, a prime buyer of Australian resource exports.
If the 1992 trough is considered as a "double dip" of the 1991 trough, it started from the peak reached on 22 May 1992, day 341 of the 1991 recovery. In today's timing terms the equivalent high would be in about 6 weeks from now. As can be seen from the graphs, the recoveries vary widely in amount and duration so this is not a forecast.
See the recent post comparing recoveries for more information on the corrections that have occurred during the recovery phase.
Hat tip to Doug Short for his more sophisticated series of charts showing Dow falls and recoveries:
http://dshort.com/charts/bear-recoveries.html?current-bear
Thursday, May 27, 2010
Comparison of Australian Stockmarket Recoveries since 1984
Yahoo has Stockmarket data for the Australian All Ordinaries since 1984. I have used that data to compare stock market recoveries from market bottoms since the deep, short, dramatic crash of 1987.
Background
On fundamentals Australia did not have a housing collapse or oversupply, had no or a very short shallow recession, has little in the way of unprovisioned bank credit losses and has relatively small unemployment, largely due to stimulus which flowed to the firms and workers in the construction industry. There is high private debt to GDP, but extremely low Government debt to GDP. The deficit is moderate compared to developed other countries. Australia issues its own currency. Some say Australia still has a housing bubble. In 2009 Australia had one of the best performing stock markets in the world in both USD and EUR terms.
Where we are - 310 of 350 days
First 350 days of Australian All Ordinaries Recoveries - Click for large image
The current (bright blue) recovery was much faster for 160 days but is now back to average (bright orange) values. Only the recovery from 2003 was higher after 310 days, so this recovery is still ahead of 4 out of 5 other recoveries.
Where we are going - looking to 1200 days
Australian All Ordinaries Recoveries - First 1200 Days - Click for large image.
The Average recovery falls away from +38% at day 310 and does not recover to current levels of +40%) until about day 450. The Average recovery does not break up from +40% until about day 560.
The Average balances the 1987 recovery which fell back to only +4.6% at day 800 (the 1991 low) against the 2003 and 1991 recoveries which each reached +94% at day 780. Other than 2003, all recoveries shown went to a significantly lower point by between day 350 and day 460.
In addition to the 1987 recovery major bottom at day 800 (1991 low) (described above):
* 1987 fell from +44% at day 186 to +24% at day 356 (14% fall) 1989 low)
* 1991 fell from +40% at day 210 to only +13% at day 468 (19.5% fall) (1992 low)
* 1992 fell from +71% at day 310 (same as where we are now) to +35% at day 529 ( 21% fall) (1995 low)
* 1995 had a correction from +26% at day 273 to +15% at day 364 (8.7% fall) (1996 low)
* 2003 eventually had a small correction from +58 at day 515 to plus +47 at day 554 (7% fall).
The outlook for the period out another 200 days to day 510 is for a real risk of a major (new or continuing) correction in the Australian broad indices.
The current downturn has been very sharp compared to other major downturns.
Short Term Action
I am, after considering all of the above, watching the 10 and 15 day simple moving averages for a buy signal as There may well be medium term (within the context of 2 to 6 year cycles) rallies before any new, if any, low is reached.
Other Major Markets
Most other major stock markets are below their 200 day simple moving averages. This is generally regarded as an indicator that a major downtrend has commenced, more so if it is on a month end or if it is held for say 5 to 10 trading days. However in a volatile mainly cross trending market it is unreliable, leading to expensive whipsawing as in the Australian All Ordinaries from 1998 to 2003.
Acknowledgements:
* Doug Short for his similar charts on the US market http://dshort.com/articles/2010/sixteen-dow-recoveries.html
* John Hussman for his current newsletter about Aunt Minnie and the risks to the market in the near to medium term. http://www.hussmanfunds.com/wmc/wmc100524.htm
Disclosure:
Multi national bond fund about 80%, mix of share funds about 20%; real estate
Currency exposure (not including real estate): 80% AUD, 20% other currency denominated.
Themes for seeking Alpha: Australia, Global markets Stocks: EWA, KROO
This article also published in part on Seeking Alpha:
http://seekingalpha.com/instablog/508569-explorer/73410-comparison-of-australian-stockmarket-recoveries-since-1984
Seeking Alpha articles on Australia:
http://seekingalpha.com/instablog/tag/Australia
http://seekingalpha.com/tag/australiastockblog
Background
On fundamentals Australia did not have a housing collapse or oversupply, had no or a very short shallow recession, has little in the way of unprovisioned bank credit losses and has relatively small unemployment, largely due to stimulus which flowed to the firms and workers in the construction industry. There is high private debt to GDP, but extremely low Government debt to GDP. The deficit is moderate compared to developed other countries. Australia issues its own currency. Some say Australia still has a housing bubble. In 2009 Australia had one of the best performing stock markets in the world in both USD and EUR terms.
Where we are - 310 of 350 days
First 350 days of Australian All Ordinaries Recoveries - Click for large image
The current (bright blue) recovery was much faster for 160 days but is now back to average (bright orange) values. Only the recovery from 2003 was higher after 310 days, so this recovery is still ahead of 4 out of 5 other recoveries.
Where we are going - looking to 1200 days
Australian All Ordinaries Recoveries - First 1200 Days - Click for large image.
The Average recovery falls away from +38% at day 310 and does not recover to current levels of +40%) until about day 450. The Average recovery does not break up from +40% until about day 560.
The Average balances the 1987 recovery which fell back to only +4.6% at day 800 (the 1991 low) against the 2003 and 1991 recoveries which each reached +94% at day 780. Other than 2003, all recoveries shown went to a significantly lower point by between day 350 and day 460.
In addition to the 1987 recovery major bottom at day 800 (1991 low) (described above):
* 1987 fell from +44% at day 186 to +24% at day 356 (14% fall) 1989 low)
* 1991 fell from +40% at day 210 to only +13% at day 468 (19.5% fall) (1992 low)
* 1992 fell from +71% at day 310 (same as where we are now) to +35% at day 529 ( 21% fall) (1995 low)
* 1995 had a correction from +26% at day 273 to +15% at day 364 (8.7% fall) (1996 low)
* 2003 eventually had a small correction from +58 at day 515 to plus +47 at day 554 (7% fall).
The outlook for the period out another 200 days to day 510 is for a real risk of a major (new or continuing) correction in the Australian broad indices.
The current downturn has been very sharp compared to other major downturns.
Short Term Action
I am, after considering all of the above, watching the 10 and 15 day simple moving averages for a buy signal as There may well be medium term (within the context of 2 to 6 year cycles) rallies before any new, if any, low is reached.
Other Major Markets
Most other major stock markets are below their 200 day simple moving averages. This is generally regarded as an indicator that a major downtrend has commenced, more so if it is on a month end or if it is held for say 5 to 10 trading days. However in a volatile mainly cross trending market it is unreliable, leading to expensive whipsawing as in the Australian All Ordinaries from 1998 to 2003.
Acknowledgements:
* Doug Short for his similar charts on the US market http://dshort.com/articles/2010/sixteen-dow-recoveries.html
* John Hussman for his current newsletter about Aunt Minnie and the risks to the market in the near to medium term. http://www.hussmanfunds.com/wmc/wmc100524.htm
Disclosure:
Multi national bond fund about 80%, mix of share funds about 20%; real estate
Currency exposure (not including real estate): 80% AUD, 20% other currency denominated.
Themes for seeking Alpha: Australia, Global markets Stocks: EWA, KROO
This article also published in part on Seeking Alpha:
http://seekingalpha.com/instablog/508569-explorer/73410-comparison-of-australian-stockmarket-recoveries-since-1984
Seeking Alpha articles on Australia:
http://seekingalpha.com/instablog/tag/Australia
http://seekingalpha.com/tag/australiastockblog
Subscribe to:
Posts (Atom)